Djibouti - Highway Maintenance Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 9111 PROJECT COMPLETION REPORT REPUBLIC OF DJIBOUTI HIGHWAY MAINTENANCE PROJECT (CREDIT 1303-DJI) SEPTEMBER 19, 1990 e Operations Division and Indian Ocean Department restricted distribution and may be used by recipients only in the performance of Its contents may not otherwise be disclosed without World Bank authorization.  DJIBOUTI HIGHWAY MAINTENANCE PROJECT (Credit 1303-DJI) PROJECT COMPLETION REPORT CURRENCY EQUIVALENTS Currency Unit = Djibouti Franc (DF) US$1 = DF 175.0000 (appraisal) US$1 = SDR 0.8906 (appraisal) US$1 = DF 177.6300 (average) US$1 = SDR 0.9130 (average) ABBREVIATIONS AND ACRONYMS ERR = Economic Rate of Return FAC = Fonds d'Aide et de Cooperation (France) MI = Ministry of the Interior MPW = Ministry of Public Works, Urban Affairs and Housing PCR = Project Completion Report PWD = Department of Public Works RN = Route Nationale (National Road) SAR = Staff Appraisal Report VOC = Vehicle Operating Cost vpd = Vehicle per day FISCAL YEAR January 1 - December 31  FOR 0MCIUL UsE ONLY THE WORLD BANK Washtngton. D.C 20433 U.S.A. Oic.e i Ocwt.qeittai Opetatmn Etva&ktNm September 12, 1990 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Project Completion Report on DJIBOUTI: Highway Maintenance Project (Credit 1303-DJI) Attached, for information, is a copy of a report entitled "Project Completion Report on Djibouti: Highway Maintenance Project (Credit 1303-DJI)" prepared by the Africa Regional Office with Part II of the report contributed by the Borrower. No audit of this project has been made by the Operaiona Evaluation Department at this time. Attachment This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  FOR OFFICIAL USE ONLY DJIBOUTI HIGHWAY MAINTENANCE PROJECT (Credit 1303-DJI) PROJECT COMPLETION REPORT Table of Contents Pate N Preface .............................................. i Evaluation Summary ............................................ 11 PART I: PROJECT REVIEW FROM IDA'S PERSPECTIVE Project Identity ................................................. 1 Background. 1 Project Objectives and Description. 3 Project Design and Organization. 3 Project Implementation. 4 Delays before Effectiveness. 4 Maintenance of the Paved Network. 4 Improvement and Maintenance of the Unpaved Network. 5 Procurement of Road Equipment. 6 Construction of workshops. 6 Training. 6 Project Risks. 7 Project Costs ............. 7 Project Results. 8 Economic Rate of Return. 9 Project Sustainability .10 IDA Performance .10 Borrower Performance .11 Project Relationships .................... I . 11 Consulting Services .11 Project Documentation and Data .11 PART II PROJECT FROM BORROWER'S PERSPECTIVE .12 PART III STATISTICAL INFORMATION Related IDA Credits .14 Project Timetable .15 Credit Disbursements .16 Project Implementation .18 Project Cost and Financing .20 Status of Covenants .22 Use of IDA's Resources .23 Annex 1: Traffic and VOC .25 Annex 2: Economic Evaluations .27 Map IBRD No. 22585 L This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  - i - DJIBOUTI HIGHWAY MAINTENANCE PROJECT (CREDIT 1303-DJI) PROJECT COMPLETION REPORT PREFACE This is the Project Completion Report (PCR) for the Highway Maintenance Project in Djibouti, for which Credit 1303-DJI in the amount of SDR 5.7 million was approved on November 30, 1982. The credit was closed on June 30, 1987, six months behind schedule. It was not fully disbursed, because SDR 38,207.47 was cancelled on January 25, 1988, when the last disbursement was made. The Infrastructure Operations Division, South-Central and Indian Ocean Department, of the Africa Regional Office prepared the Preface, Evaluation Summary, Parts I and III. IDA requested the Government to prepare a PCR several times, but it did not receive any reply. On May 11, 1990, IDA sent Part I and III of this PCR to the Government for comments which were given in a letter dated July 17, 1990. The content of the letter is given in Part II. Preparation of this PCR was started during IDA's last supervision mission of the project in October 1986. The PCR is based, inter alia, on the Staff Appraisal Report, Credit Agreement, supervision reports, correspondence between IDA and the Government, and IDA's internal memoranda.  - iii - DJIBOUTI HIGHWAY MAINTENANCE PROJECT (CREDIT 1303-DJI) PROJECT COMPLETION REPORT EVALUATION SUMMARY Objlectives The Highway Maintenance Project, IDA's first involvement in Djibouti's transport sector, aimed at assisting the Government in strengthening its road organization to carry out effective, economic and efficient road improvement and maintenance operations, thereby providing easier and more reliable access to the hinterland and reducing transport costs (Part I - para 1.11). Implementation Experience The achievements for periodic maintenance of paved roads largely exceeded the original objective, e.g. 66 km. vs. 25 km. (Part I - para. 1.16). Spot improvements of about 800 km of priority earth roads and tracks by labor intensive techniques proved to be more difficult than expected. The Department of Public Works (PWD) improved about 85 km of priority roads to gr. vel standards using mechanized units i.lstead (Part I - paras 1.17-1.20). Procurement of the equipment financed by IDA was delayed because of the time required for international competitive bidding, Government's lack of familiarity with IDA's policies and procedures, and IDA's slowness to review bidding documents (Part I - para 1.21). Results The project fulfilled the objectives of carrying out road improvement and periodic maintenance of paved roads. However, neither the maintenance of gravel and earth roads, nor the strengthening of the road organization were carried out as expected (Part I - paras 1.32-1.34). The quality of all work executed under the project was good. Project measurable targets were all attained except for those set for the routine maintenance of earth roads and tracks (Part I - para 1.32). Measurable targets for training were not set at appraisal. Also, no information on the number of PWD's local employees trained under the project and results obtained is available, with the exception of the chief mechanics (Part I - paras 1.33). The Ministry of Public Works (MPW) and PWD were reorganized under the project. Structurally, the new organization was sound, even if it required to be further improved. The strengthening of the road organization to carry out effective, economic and efficient road improvement and maintenance operations was not fully realized (Part I - para 1.34).  - iv - Sustainability It is expected that throughout its economic life the project will maintain at least the same level of net benefits as estimated in this PCR on very conservative assumptions (Part III - Annex 2). However, benefits would drop considerably if (i) the Government should not make available sufficient funds to maintain the investments made under the project adequately, and (ii) no technical assistance should be made available to PWD (Part I - paras 1.39 and 1.40). Findings and Lessons Learned The main lesson learned concerns the progress that can actually be made in strengthening organizations and in training local staff. The project confirmed what had already been learned from many other similar projects. Institution building and training require continuously sustained effort over a long period of time before tangible and durable results can be obtained (paras 1.33 and 1.34).  DJIBOUTI HIGHWAY MAINTENANCE PROJECT (Credit 1303-DJI) PART I: PROJECT REVIEW FROM BANK'S PERSPECTIVE A. Proiect Identity 1.01 Project Data Project Name Highway Maintenance Project Loan Number 1303-DJI Country Department AF3 Country Republic of Djibouti Sector Transport Subsector Highways Government's Executing Agency Ministry of Public Works (MPW) through its Public Works Department (PWD) Follow-on Project None B. Backzround Macro Policy Objectives 1.02 Djibouti, which became independent in 1977 and a member of the Bank and its affiliates in October 1980, is a small country of about 22,000 km2. Most of its population of about 400,000 inhabitants lives in Djibouti City, capital as well as port and railway terminal, where most of the economic activities are concentrated. The hinterland, characterized by high mountains and plateaux, is generally arid and hot and inhabited by poor pastoral population. 1.03 Djibouti's natural resources are very limited. In particular, agriculture is almost non-existent and livestock scanty because of the extreme shortage of arable and pasture land and water; the productive sectors, therefore, have plaid only a minor role. For these reasons and because of Djibouti's location, economy has, therefore, been very much oriented to financial, commercial and shipping services for the region. 1.04 To support its role as a regional center, Djibouti's development policy was mainly concentrated on the modernization of its infrastructure. Because of population pressures in the city and an impending shortage of potable water there, as well as strategic reasons, the Government began encouraging development in the hinterland. Ac the beginning of the '80s, the Government redefined the country's economic goals. Though Djibouti's economy would remain oriented to services, priority was given to reducing poverty in urban and pastoral areas, to decrease food and energy dependency and to promote agriculture and industrialization.  Sector Development Objectives 1.05 In 1982, the transport system comprised about 2,800 km of roads and tracks 1, 110 km of railway , the Djibouti port and an international airport and several airstrips. In particular, the road network consisted of 300 km of paved roads and streets and some 2500 km of earth roads and tracks. 1.06 The paved network included the Djibouti-Ethiopian border road (217 km) and 83 km of streets, most of them in Djibouti City. The Djibouti- Ethiopian border road was in fair condition, but the first section (30 km), built in 1953, needed to be rehabilitated, and the middle section, the Grand Bara (25 km) needed resurfacing. Most of the tracks were in poor condition and passable only by 4-wheel drive vehicles. 1.07 On June 2, 1983, the Government established, by decree, a National Road Network of about 1,100 km, including 300 km of paved roads and streets and about 800 km of priority earth roads and tracks, which would have first call on the limited funds available for road improvement and maintenance, and for which responsibility would be of MPW. 1.08 Before Djibouti's independence in 1977 and soon after, the Government's main objective for the transport sector had been to support Djibouti's role as a regional transhipment point. Most of the investments, therefore, were made in the port of Djibouti, as well as in the railway and road from Djibouti City to the Ethiopian border. 1.09 The development of the rest of the road network, consisting of tracks, had been neglected, and the outlying areas, where the poor pastoral population lived, had virtually been cut off from Djibouti City, economic and political capital. The Government, then, changed its policy and began investing in the development of the road network to supply those areas and unify the country. 1.10 Because the interior of Djibouti was rather sparsely populated, so that, in general, building new roads would have low economic rate of return, maintenance and improvement of existing earth roads was essential to maintain liaison between the capital and other settlements. 1 The figures given in the Project Completion Report are those of the Appraisal Report. They do not match, however, those given in the consultants' study (Etude diagnostique de l'entretien routier). According to this study the length of the road network was 2,906 km, of which 1,106 km were national roads, following the Government's decree of April 1982, and 1,800 km were district tracks. The national road network comprised 249 km of paved roads including the Djibouti City-Ethiopian border (219 kIn), 32 km of gravel road and 825 km of tracks. The 68 km of streets, of which 50 km in Djibouti, were not part of the national network.  E. Proiect Implementation Delays before Effectiveness 1.14 The project was scheduled to be presented to the Board on September 21, 1982 but the Government sent IDA its agreement to the negotiated documents only at the end of October and did not comply with one condition 3 for Board presentation. To expedite the project, IDA had to change that condition into a condition of effectiveness. The project was presented to the Board on November 30, 1982, and the Credit Agreement signed on January 6, 1983. 1.15 The effectiveness date had to be postponed three times because of the Government's lack of familiarity with IDA's requirements for effectiveness. The Credit became finally effective on October 26, 1983, almost one year after the Board presentation. Project implementation began, therefore, about one year and a half later than scheduled at appraisal. Maintenance of the Paved Network 1.16 Originally, the project provided for periodic maintenance of the paved network consisting of resurfacing 25 km of the Grand Bara section of RN 1 (Djibouti City-Ethiopian border road, 219 km). Neither the Appraisal report nor the Consultants's study (Etude diagnostique de l'entretien routier) specified whether the resurfacing would consist of a single or double surface treatment. Under the project, PWD actually carried out the following periodic maintenance work: (a) 25 km of single surface treatment on RN 1 (Djibrp--i City- Ethiopian border, 219 km), (b) 8 km of double surface treatment on RN 3 (Balbala-Khor Ambado, 23 km) (c) 10 km of double surface treatment on RN 5 (Grand Bara-HolHol, 64 km), (d) 7 km of double surface treatment in Djibouti City (new streets) (e) 16 km of single surface treatment in Djibouti City (existing streets). 3 By the decree of June 16, 1982, the Government had given MPW responsibility for improving and maintaining the National Road Network, including the streets of Djibouti City. A condition of Board presentation first, and then a condition of effectiveness, was that the Government promulgated a second decree officially establishing the organization of MPW and defining its functions, including routine maintenance of the streets drainage and sidewalks of Djibouti City. This also involved the transfer of about 500 laborers ard related operating funds from the Ministry of the Interior to the Ministry of MPW.  - 3 - C. ProJect Obiectives and DescriRtio Prolect Obiectives 1.11 In accordance with the Government's policy, the Highway Maintenance Project, IDA's first involvement in Djibouti's transport sector, aimed at assisting the Government in strengthening its road organization to carry out effective, economic and efficient road improvement and maintenance operations, thereby providing easier and more reliable access to the hinterland and reducing transport costs. Project Components 1.12 These were the Project components: (a) a three-year road improvement and maintenance program, including resurfacing of about 25 km of paved roads, routine maintenance on 300 km of paved roads and streets, and spot improvements and maintenance with hand tools on about 800 km of priority earth roads and tracks, (b) procurement of equipment, spare parts, tools, materials and supplies for the works under (a), (c) construction of a workshop in Djibouti City and of a service station in Dikhil, and expansion of the service station in Tadjourah, (d) procurement of equipment and tools for workshops, and (e) consulting services comprising (i) technical assistance to strengthen the road or anization, train personnel and do preinvestment studies to identify the improvement needs and priorities of the road network, and (ii) fellowships for the training of road maintenance personnel. D. Proiect Design and Organization 1.13 The project design and scope resulted from the findings of IDA's identification mission and a preinvestment study carried out by consultants. The desi,n was very simple, not innovative and represented a realistic approac to the road-development needs of the country. Only one Government agency, with the assistance of consultants, was responsible for the project. 2 These studies included detailed engineering for the improvement to gravel standards of about 250 km of tracks and preinvestment studies for the rehabilitation of the first section (30 km) of the paved road from Djibouti City to the Ethiopian border (RN 1).  Namely, about 66 km of surface treatment, 41 km (164%) more than planned. Also, PWD carried out the routine maintenance of about 300 km of paved roads and streets as provided for under the project, Improvement and Maintenance of the Unpaved Network 1.17 The project provided for spot improvements and maintenance with hand tools of about 800 km of priority earth roads and tracks. This work was to be carried out mainly by labor intensive techniques, but this proved to be even more difficult than expected 4. In 1984, PWD set up a firs-z gang and put it to work on RN 3, but it advanced so slowly mainly because of the rocky soil that PWD decided to replace it with a mechanized unit which had been set up with existing equipment. 1.18 In September 1984, PWD set up a second mechanized unit with equipment procured under Japan's parallel financing and, one year later, in September 1985, it set up a third mechanized unit with the first batch of equipment procured under IDA's financing. Actually, therefore, PWD did not carry out any spot improvement by labor intensive techniques but it improved the following priority roads to gravel standards using mechanized units: km (a) RN 3, Dorale-Khor Ambada 7 (b) RN 5, Ali Sabich-Ali Adde 7 (c) RN 6, Dikhil-As Eyla 37 (d) RN 14, Tadjourah-Obock 34 5 Total 85 1.19 Neither IDA's files nor the consultants' final report give any information about whether the 800 km of the earth priority road and tracks received any routine maintenance at all. The Ministry of Interior (MI) was to transfer about 500 laborers and the related operating funds to MPW, but it never did it. As most of MI's laborers were old, unskilled and not prone to work in the field under severe conditions, and their transfer was politically difficult, MPW once proposed to set up a road fund, fed by taxes on fuel, to finance road improvement and maintenance. But the proposal never came through. 1.20 MI is the only agency which might have done some routine maintenance 4 Excerpt from the Project Brief: "In view of the heavy unemployment in Djibouti, it would appear desirable to maximize the use of labor-intensive techniques. However, while every attempt will be made to use labor-intensive methods, this may prove difficult, because (i) many parts of the country are sparsely populated so that workers would have to be brought considerable distances from the town or from refugee camps, (ii) it may not be easy to find people prepared to do labor-intensive road maintenance work in the extreme conditions of the Djibouti hinterland (where temperature above 50 degrees centigrades are common) and (c) there may be problems of financial justification due to the relatively high minimum wage in Djibouti." 5 Including the by-pass to Ribta  on those roads using the labor force and the funds which it should have transferred to MPW. Hodever, this is verv unlikely, and if MI did carry out some routine maintenance, the results must have been very poor 6, Procurement of road equipment, etc (see 1.12 (b)) 1.21 The equipment financed by Japan was procured expeditiously; by September 1984, it had been all delivered and most of it was operational. Procurement of the equipment financed by IDA was delayed because of the time required for international competitive bidding, _overnment's lack of familiarity with IDA's policies and procedures, and IDA's slowness to review bidding documents. By September 1985, the first batch had been delivered and was operational. The second and third batches were delivered at the beginning of 1987, the last year of project execution. ConsLruction of Workshops 1.22 The project provided for the construction of workshops in Djibouti City and service stations in Tadjourah and Dikhil. All the work, carried out by contract, was completed. However, the construction of the Djibouti workshop progressed very slowly, and the Government had to replace the contractor. As a result, the new Djibouti workshop could not be used under the project, because the work was finished at the beginning of 1987, i.e. when the improvement and maintenance work was completed. Eventually, the Djibouti workshop became operational and supported PWD maintenance operations. Training 1.23 The project provided for (i) on-the-job training of counterparts, including engineers, workshop mechanics, road technicians, store keepers and accountants, and (ii) overseas fellowships for eight instructors who would train equipment operators and road technicians. Training of counterparts and assignment of fellowships, however, proved to be difficult for lack of suitable candidates. 1.24 Three mechanics only availed themselves of the fellowships by attending training session in France. The Government, then, with IDA agreement, requested the consultants to hire an additional expert, a mechanic, to train mechanics on-the-job for ten months. This program was carried out from January to September 1986, during which 16 chief mechanics were trained. 1.25 Before and after the training, the expert tested the trainees' 6 Excerpt from SAR: - "However, as the municipalities and MI lack the funds, expertise and equipment to carry out these tasks, PWD is regularly called upon to provide these services on an emergency basis (paragraph 1.17, page 5)." and - "Teams of MI's laborers nominally in charge of routine maintenance Df earth roads and tracks, produce no tangible results (paragraph 2.03, page 9).".  techLnical knowledge. The outcome indicated that, on the average, the trainees' knowledge had improved but that there was still a long way to go in order to obtain an acceptable level of technical knowledge, i.e. training should continue. 1.26 The expert came up against several difficulties during the training. One was the late completion of the Djibouti workshop and another,inherent to on-the-job training, was the too short time available to explain the methods to find out the causes of the breakdowns, because of the urgency to repair the equipment. Proiect Risks 1.27 SAR (para 3.08) pointed out that the main risks of the project were the usual risks associated with road maintenance projects. Project implementation might be slowed down by delays in procurement of equipment, materials and supplies, recruitment of technical assistance, construction of workshops and provision of local funds, as well as by productivity lower than the modest levels assumed. These risks were compounded by the fact that the project was IDA's first operation in the road subsector in Djibouti. 1.28 Actually, project implementation was delayed because of difficulties in the (i) preparation and approval of bidding documents for procurement of equipment, spare parts, materials and supplies and (ii) construction of workshops (paras 1.21 and 1.22). However, project implementation was delayed mainly before the actual implementation of the project started, because of delayed credit effectiveness (para 1.15). Productivity was not one of t,'e factors that delayed project implementation. 1.29 Also, SAR pointed out that, because the project was IDA's first operation in the road subsector in Djibouti, more frequent supervision would also be required in the initial project stages. Looking at the number and dates of the actual supervision missions, it does not seem that IDA met this requirement adequately. 1.30 The Government and consultants were actually faced with another difficulty not mentioned among the risks in SAR, but identified since project preparation, namely the difficulty for PWD in keeping and recruiting trained and experienced staff at all levels 7. This affected the training program adversely (para 1.23). Project Costs 1.31 The estimated total cost at appraisal was US$8.0 million, excluding taxes, and the estimated foreign cost was US$6.4 million entirely financed by IDA. The actual total cost, net of taxes, turned out to be much higher, i.e. 7 The Project Brief pointed out that "this difficulty is aggravated by the proximity of Saudi Arabia, which needs, and therefore imports, qualified staff. The project will have to make a major training effort, even if many of the people trained may not stay in their jobs, because they may seek more remunerative employment, or more pleasant working conditions elsewhere.".  - 8 - US$13.9 million of which the foreign cost was US$10.7 million. An increase of US$5.8 million (72%) due to procurement of more equipment, financed by Japan, as well as more materials and supplies, financed by the Government, than estimated at appraisal 8. Breakdown of the project financing in US$ million was as follows: SAR Actual Local Foreign Total Local Foreign Total IDA - 6.4 6.4 (80%) - 6.2 6.2 (45%) Government 1.6 - 1.6 (20%) 3.3 2.8 6.1 (44%) Japan - - - ( 0%) - 1.6 1.6 (11%) Total 1.6 6.4 8.0 3.3 10.6 13.9 F. Project Results 1.32 The quality of all work executed under the project was good. The productivity of the mechanized units for both resurfacing of paved roads and improvement to gravel standards of tracks, was higher than expected. Project measurable targets were all attained except for those set for the routine maintenance of earth roads and tracks which was not carried out under the project (para 1.19). Actually, the targets for resurfacing and routine maintenance of paved roads and for improvement to gravel standards of earth roads and tracks, were exceeded. 1.33 Measurable targets for training were not set. Also, no information on the number of PWD's local employees trained tnder the project and results obtained is available, with the exception of the chief mechanics (paras 1.23- 1.25). T.e only replacement of expatriates with Djouboutians, mentioned in either IDA's files or the consultants' final report, is that of PWD's director in July 1986. The consultants estimated in their report that PWD's expatriate staff would decrease from 13 9 in 1986 to 12 in 1987, to 6 in 1988 and to 5 in 1989. 1.34 MPW and PWD were reorganized in 1984. PWD split up into two departments, Department of Public Works and Department of Town Planning and Housing. Structurally, the new organization was sound even if it required to be further improved. However, there was still lack of qualified staff at all levels but clerical, and a need for a lot more of training. PWD still needed many expatriates (para 1.33) to run adequately and properly. Because at the end of the project, MPW, and in particular PWD, still depended heavily on technical assistance, the strengthening of the road organization to carry out effective, economic and efficient road improvement and maintenance operations 8 Fluctuations of the rates of exchange between the US dollar and the Special Drawing Rights and between the US dollar and the Djibouti franc were so small that they did not have any appreciable impact on the actual project cost. At appraisal, SDR 1=US$1.123 and US$1=DF 175. The average rates of exchange for the project period were: SDR 1=US$1.095 and US$1=DF 177.63. 9 According to the appraisal report, there were 26 expatriates in PWD in 1982, provided by Fonds d'Aide et de Cooperation (FAC). According to the consultants' final report, there were 13 at the end of 1986.  -9- cannot be considered as realized. Economic Rate of Return (ERR) 1.35 At appraisal, because no historical data on traffic volumes existed 10, and the economic activity outside Djibouti City was limited, a conservative traffic growth of 2.5% p.a. for paved roads and 1% p.a. for earth roads and tracks were assumed. Vehicles operating costs (VOC) and savings were derived using the Highway Design and Maintenance Standards Model (HDM). Not all data used at appraisal to evaluate ERR, however, are on file, namely VOC for paved roads. ERR of the resurfacing of 25 km of paved road was evaluated at 58% on the basis of an economic life of 5 years, and that of the routine maintenance of both paved and unpaved roads was evaluated at more than 100%, on the basis of an economic life of 3 years. 1.36 The Government did not prepare a completion report for the project, in spite of IDA's numerous requests. IDA's last supervision mission visited Djibouti in October 1986, and because there has been no follow-on project, no other transport mission has visited Djibouti so far. The consultants' final report does not deal with economic re-evaluation of the project. PWD carried out two traffic counts in 1983 and 1984, i.e. before the start of project implementation, but they were taken only on RN 1, and too close to Djibouti City for the results to be of any use. Consequently, no data, except for actual costs, were available to re-evaluate ERR. 1.37 In spite of this lack of data, ERR was re-evaluated all the same, but on very conservative assumptions (Part III - Annex I) as follows: (a) The economic re-evaluation of the resurfacing (single surface treatment) of 25 km of RN1 (para 1.16 above) was done using the same traffic, 60 vpd, as at appraisal, but assuming a 0% growth p.a. and no savings in reconstruction cost after 5 years, as assumed at appraisal 11. Also, since no VOC for paved roads were on file, it was assumed that VOC of a representative vehicle for a good paved road was about 13% only lower than that given for good gravel roads in the 1981 study: "Etude diagnostique de l'entretien routier". On these assumptions, ERR was about 16% (Part III - Annex 2, para 1), whereas at appraisal, it had been evaluated at 58%. (b) An economic evaluation of the resurfacing (double surface treatment) of 18 km of RN3 and RN5 (para 1.16 above), not originally included in the project, was carried out on the same conservative assumptions. ERR was about 10% (Part III - Annex 2, para 2). ERR turned out to be lower than that of RN1 because of the higher cost (double surface treatment instead of single surface treatment). (c) The economic evaluation of the resurfacing of 23 km of streets in Djibouti City, not originally included in the project, showed an ERR of about 25% (Part III - Annex 2, para 3). (d) An economic re-evaluation of the routine maintenance of about 300 km of paved roads and streets (para 1.16 above) was also carried out 10 The only traffic data available were those from a one-week count made on selected roads and tracks by consultants in 1981. 11 Savings in reconstruction cost after only five years were not considered also because the 25-km-long road section actually required only a single surface treatment, hence it must have been in better condition than assumed at appraisal.  - 10 - on the same conservative assumptions. ERR was about 57% (Part III - Annex 4, para 3) versus an estimate of 100% at appraisal. Routine maintenance of 800 km of tracks was not carried out under the project. It was replaced by the improvement to gravel standards of 85 km of tracks (para 1.18 above). (e) The economic evaluation of this %mprovement to gravel standards of 85 km of priority tracks was done usiag a traffic of only 10 vpd, i.e 50% of the highest traffic estimated at appraisal for this type of roads, and assuming a growth p.a. of 0%. VOC were assumed to be the same as those given in the 1981 study: "Etude Diagnostique de 1' Entretien Routier". On these assumptions, ERR was about 14% (Part III - Annex 2, para 5). 1.38 The overall project rate of return was estimated at 22% as compared to more than 100% estimated at appraisal, largely because of the very conservative assumption on traffic growths which were utilized for the PCR, in the absence of adequate data. G. Proiect Sustainability 1.39 The work executed under the project was of good quality, the Government provided more funds than expected (44% of project total cost as against 20% estimated at appraisal) showing its determination to improve and maintain the road network, and PWD with technical assistance performed relatively well under the project. It is expected, therefore, that throughout its economic life the project will maintain at least the same level of net benefits estimated, on very conservative assumptions, in this PCR as shown in Part III - Annex 2. 1.40 However, benefits might drop considerably if (i) the Government should not make available sufficient funds to maintain the investments made under the project adequately, and (ii) no technical assistance should be made available to PWD. As IDA has not decided to finance a follow-on project in the transport sector as yet, it is not in a position to address these issues in detail. H. IDA Performance 1.41 The project was small and simple without any complicated transport issues and not innovative for IDA. It was delayed mainly because of the difficulties that the Government encountered in meeting conditions for credit effectiveness, but then it was executed smoothly without major issues. IDA might have been more speedy in reviewing bidding documents and could have supervised the project more frequently, since the project was IDA's first involvement in the highway subsector, but probably it could not because of staff constraints.  - 11 - I. Borrower Performance 1.42 The Government did not perform well during the preparation of the project, namely in meeting conditions of credit effectiveness. The project, however, was IDA's second operation in Djibouti 12, and the Government's lack of familiarity with IDA's requirements was the main reason for this poor performance, that delayed project implementation by about one year and a half. During project execution, the Government performed satisfactorily. The Borrower's main weaknesses were its unfamiliarity with IDA's procedures and requirements (which IDA should have addressed through increased supervision) as well as its inability to supply qualified staff to the project. Both were well known by IDA during project preparation and can hardly be imputed to bad performance. J. Proiect Relationships 1.43 The Borrower-IDA relationship was generally good. At a given time, however, there were conflicting opinions about the responsibility for the delays in finalizing bidding documents. PWD accused IDS of not reviewing bidding documents in a speedy manner, and IDA thought that the Government did not prepare bidding documents as it should, i.e. according to IDA's requirements. K. Consulting Services 1.44 The consultants performed well. Their reports, in particular the final one, however, could have been a little more accurate and clear. L. Prolect Documentation and Data 1.45 The legal documents and the appraisal report were prepared, as usual, with care and are clear. However, as the Government was not familiar with IDA's procedures and requirements (paras 1.21 and 1.43), occasionally, it had problems in interpreting and applying some of the credit agreement conditionalities. Also, not all data used for ERR evaluation are available in the appraisal report or any other document in the files. 12 IDA's first operation in Djibouti was a US$3 million credit for a technical assistance project approved by the Board on April 20, 1982.  - 12 - PART II. PROJECT REVIEW FROM BORROWER'S PERSPECTIVE 2.01 Parts I and III were sent to the Borrower on May 11, 1990 with the request to provide Part II. The following response to that request was received from the Director of Public Works in his letter dated July 17, 1990. 2.02 "In response to your letter dated May 11, 1990, please find hereunder the comments from the Borrower: (i) Concerning Part III of the PCR, I confirm that the information it contains is accurate. (ii) Concerning Part I, the Government has the following comments: (a) The program of improvement of gravel roads by mechanized brigades is being continued. Since the beginning of the Highway Maintenance Project, which had identified a priority network (of tracks) of 250 km, 186 km of tracks have been improved and it is intended to complete the program by 1994. (b) The strengthening of the road organization has been achieved as projected at the level of the subdirectorates of Dikhil and Tadjourah with the commissioning of routine maintenance crews (for pavement, shoulders, drainage and structures), as well as the setting up of specialized gangs (for heavy equipment and engines) in the equipment yards. (c) In addition to this, the training organization developed (under the project) has been maintained and is still operating, especially for operation and maintenance of heavy equipment and vehicles (one full time trainer since 1987). We are currently envisaging the setting up of a training cycle for road maintenance site foremen in 1991. Moreover, since 1988, training sessions for head foremen and topographers have been offered in the Vocational Training Center for Building and Public Works with technical and financial assistance from the French Cooperation. (d) MPW's suggestion (to the Government) to create a road fund to collect a fuel tax earmarked for the upgrading and maintenance of National Roads has not been adopted. (e) The setting up of a Road Planning Unit by 1993 is a major objective of DPW in order to analyze the road network and promote a road maintenance policy (with appropriate data processing tools). (f) A number of technical assistants (10) provided by France are still present within DPW. The replacement of these by Nationals with equivalent skills is envisaged at about 50% by 1994.  - 13 - (g) DPW does its best to maintain the road assets in good condition, those built before, during, and after the project. Routine maintenance is financed under the National Budget. Financing from various bilateral sources is expected for the periodic maintenance of RN1 as follows: Saudi Fund-30 km completed; African Development Fund-100 km, envisaged in 1991; and French Cooperation-40 km envisaged in 1992. (iii) Evaluation of Bank's performance: as this was the first project in the infrastructure sector, it is difficult (for us) to have an opinion on it. However, the project has been completed with the main objectives attained, which can be Interpreted as a good performance and capacity to adjust adequately as project implementation goes on. The flexibility shown in the Djiboutian context has been well appreciated. (iv) Evaluation of Borrower's performance: the Borrower's performance has been very positive especially after project completion. The continuation of training, the willingness to improve the road maintenance organization and management, the provision of substantial funding for routine and periodic maintenance of national paved or upgraded unpaved roads, all show a national concern (for road maintenance) following the implementation of the project. (v) Evaluation of Bank-Borrower relationship during project implementation: relationship must have been satisfactory if one considers that the project has been completed without unreasonable delay. The multiple small problems (arisen during project implementation), resulting mainly from the need to comply with Bank guidelines, might have had less impact had the supervision mission been more frequent."  - 14 - PART III 3lated Bank Credits Year of Title Purpose Approval Status Comments 1518-DJI The Project includes upgrad- 1984 under way none .ban ing and drainage of streets, refuse collection, equipment, technical support, training, management of urban invest- ments and housing credit. It is expected to help improve the living conditions of the urban poor in the city of Djibouti, and strengthen the absorptive capacity of key agencies involved in the urban sector.  - 15 - 3.02 Proiect Timetable Date Date Date Item Planned Revised Actual - Identification 13 - 03/81 - Preparation 14 04/81 _ 04/81 - Appraisal Mission 10/81 - 11/81 - Credit Negotiation 04-05/82 05/82 06/82 Board Approval 15 06/82 06/82 11/82 - Credit Signature - - 01/06/83 - Credit Effectiveness 03/31/83 06/08/83 09/15/83 10/31/83 10/26/83 - Credit Closing 12/3.1/86 - 06/30/87 16 - Project Completion 06130/86 12/31/86 12/31/87 13 The date shown herein is that of the Project Brief. However, the project was actually identified in October 1980 by an IDA's mission which also began the project preparation (see Terms of Reference of October 6, 1980 and Back- - Offic -eport of November 12, 1980). 14 In at..-rdance with the Project Brief, the date given herein is when consultants started the road maintenance study (see para 15 of the Project Brief). However, actual preparation began in October 1980 (see note 13). 15 Board presentation delayed because the Government sent IDA its agreement on the negotiated documents only on October 26, 1982 and did not meet the conditions of Board presentation. In order to present the project to the Board, IDA had to change the conditions to Board presentation in conditions to effectiveness. 16 The Credit Account was closed on Jan 25, 1988, when the last disbursement was made (see paras 3.03 A and B).  - 16 - 3.03 Credit Disbursements A. Cumulative Estimated and Actual Disbursements (US$'000 equivalent) Actual Actual as % of Appraisal as % of Credit Quarter ending Estimate Actual Estimate Amount Jun 30, 1983 200 0 0 0 September 30, 1983 600 0 0 0 December 31, 1983 1,200 0 0 0 March 31, 1984 1,600 200 13 3 June 30, 1984 2,500 400 16 6 September 30, 1984 3,000 500 17 8 December 31, 1984 3,600 700 19 11 March 31, 1985 4,200 900 21 14 June 30, 1985 4,900 1,300 27 20 September 30, 1985 5,300 2,300 43 36 December 31, 1985 5,900 3,600 61 56 March 31, 1986 6,100 4,200 69 66 June 30, 1986 6,200 4,900 79 77 September 30, 1986 6,400 5,200 81 81 December 31, 1986 5,500 86 86 March 31, 1987 5,870 92 92 June 30, 1987 5,940 93 93 September 30, 1987 5,940 93 93 December 31, 1987 6,030 94 94 January 25, 1988 6,201 97 97 Closing date: Credit 12/31/86 06/30/87 Credit Account - 01/25/88 (last disbursement)  - 17 - 3.03 Credit Disbursements (continued) B. Final Disbursement Allocation in SDR by Category Disbursements & Category Cancellations Total (1) Civil Works: Construction of Workshops 1,636,914.65 (2) Expert and Consulting Services and Fellowships 1,968,435.46 (3) Equipment, spare parts, materials, tools and supplies (except crushed stones) (a) directly imported 1,364,106.94 (b) purchased locally 692,335.48 Total Disbursements 5,661,792.53 Cancelled on January 25, 1988 38,207.47 Original Credit Amount 5,700,000.00  - 18 - 3.04 Prolect Implementation Project Targets SAR Actual Actual a0 Item Activity Targets Targets 2 of SAR - Workshops - Design and Preparation of Bidding Documents - Start date 04/01/83 02/26/84 - Completion date 06/30/83 07/12/84 - Time in months 3.0 4.5 150 - Bidding, evaluation and award - Start date 07/01/83 08/15/84 - Completion date 12/31/83 01/27/85 - Time in months 6.0 5.5 92 - Construction Work - Start date 01/01/84 03/15/85 - Completion date 17 12/31/84 01/31/87 - Time in months 12.0 22.5 188 Equipment, 18 _ Preparation of bidding Materials documents & Supplies - Start date 04/01/83 01/15/84 - Completion date 06/30/83 08/02/84 - Time in months 3.0 6.5 217 - Bidding, Evaluation and Award 19 - Start date 07/01/83 08/84 - Completion date 12/31/83 02/23/85 - Time in months 6.0 6.0 100 - Delivery 20 - Start date 01/01/84 10/15/84 - Completion date 06/30/84 06/30/87 - Time in months 6.0 32.5 542 17 The workshops in Dikhil and Tadjourah were completed on January 12, 1986 and November 19, 1985 according to the contract completion time. The Djibouti workshop, however, was completed on January 31, 1987 about eight months later than scheduled because of the contractor's defaulting. 1 The actual targets given here refer to preparation of documents for international competitive bidding. 19 The actual targets given here refer to international competitive bidding. 20 The actual targets ;.iven here refer to deliveries of goods procured through ICB, LCB and international and local shopping. Ihe bulk of equipment and spare parts, procured through ICB, was delivered from June through September 1985.  - 19 - SAR Actual Actual as Item Activity Targets Targets 2 of SA T.A. for Road - Shortlisting of Maintenance Consultants 21 - Start date 07101/82 N.A. -Completion date 09/30/82 N.A. - Time in months 3.0 - Proposals -Start date 10/01/82 N.A. - Completion date 12/31/82 N.A. - Time in months 3.0 - - Evaluation and Award -Start date 01/01/83 01/26/83 -Completion date 03131/83 01112/84 - Time in months 3.0 11.5 383 - Services -Start date 22 04/01/83 10/09/83 - Completion date 03/31/86 03/22/87 - Time in months 36.0 41.5 115 Preinvestment - Shnrtlisting of Stidies Consultants - Start date 07/01/82 07/04/83 -'Completion date 09/30/82 11/30/83 - Time in months 3.0 5.0 167 - Proposals - Start date 10/01/82 01/19/84 -Completion date 12/31/82 05/03/84 - Time in months 3.0 3.5 117 - Evaluation and Award -Start date 01/01/83 05/03/84 - Completion date 03/31/83 08/15/84 - Time in months 3.0 3.5 117 - Services - Start date 04/01/83 10/09/83 - Completion date 06/30/84 10/15/85 - Time in months 15.0 24.0 160 21 Actually, there was no short list. The Government proposed, and IDA agreed, hiring the same consultants who had carried out the Maintenance Study. 22 The Government aporoved of the contract for technical assistance on November 1, 1984 and notified the consultants of the approval on January 12, 1984, when the consultants' services should have started. However, they actually started on October 9, 1983 when, following the Government's and IDA's approval, the mission leader arrived in Djibouti  - 20 - 3.05 Project Cost and Financing A. Proiect Costs in USS'O00 Appraisal Estimate 23 Actual FQreign Foreign as S of as 2 of Pr.ject Components Local Foreign Total Total Local Foreign Total Total A. RO D IMPROVEMENT S 2 AN MAINTENANCE - >;quipment, Spare .'arts, Tools 130 1,178 1,308 90 358 3,219 3,577 90 - I!aterials and lJupplies 69 626 695 90 362 3,259 3,621 90 - Ualaries 600 0 600 0 1,838 0 1,838 0 Jubtotal 799 1,804 2,603 69 2,558 6,478 9,036 72 B. WO KSHOPS - Construction 553 2,367 2,920 81 417 1,780 2,197 81 - Design and Supervision 23 208 231 90 13 113 126 90 - Equipment, Spare i'arts, Tools 66 597 663 90 i3 481 534 90 jubtotal 642 3,172 3,814 83 41 3 2,374 2,857 83 C. CO'SULTING SERVICES AN.) FELLOWSHIPS - Technical Assistance 104 941 1,045 90 167 1,439 1,599 90 - ireinvestment Studies 46 402 448 90 41 370 411 90 - i'ellowships 9 79 88 90 2 18 20 90 .ubtotal 159 1,422 1,581 90 203 1,827 2,030 90 D. TCOAL COST 1,600 6,398 7,998 80 3,244 10,679 13,923 77 23 Including physical and price contingencies.  - 21 - 05 Proiect Cost and Financina (continued) B. Pro1ect Financins in USS'000 Appraisal Actual IDA as IDA as S of % of roj.ct Components Gov IDA Total Total Gov IDA Japan Total Total A. RO JD IMPROVEMENT & JAINTENANCE Eqdipment, Spare Pa_ts and Tools 130 1,180 1,310 90 110 1,825 1,643 3,578 51 Ma:erials anl Supplies 69 626 695 90 3,280 341 0 3,621 9 Salaries 600 0 600 0 1,838 0 0 1,838 0 Su)total 799 1,806 2,605 69 5,228 2,166 1,643 9,037 24 WO'KSHOPS Coistruction 553 2,367 2,920 81 745 1,452 0 2,197 66 Deiign and Su!ervision 23 208 231 90 0 126 0 126 100 Eqtipment, Spire Parts, Toils 66 597 663 90 106 428 0 534 80 Suutotal 642 3,172 3,814 83 851 2,006 0 2,857 70 C. COISULTANTS AND FE .LOWSHIPS Te:-hnical Assistance 104 941 1,045 90 0 1,599 0 1,599 100 Pre,investment Stuadies 46 402 448 90 0 411 0 411 100 Fellowships 9 79 88 90 0 20 0 20 100 Suitotal 159 1,422 1,581 90 0 2,030 0 2,030 100 D. TC'AL FINANCING 1,600 6,400 8,000 80 6,079 6,202 1,643 13,924 45  - 22 - 3.06 Status of Covenants Covenant Article Status The Borrower shall provide the 3.01 Complied with funds, facilities, services and other resources required for the purpose. Before granting the fellow- 3.04 Complied with partially. Onlv ships under Part C(IV) of the three employees of PWD, out of Project, the Borrower shall eight, availed themselves of establish training curricula fellowships by attending acceptable to the Association. training sessions in France. The Borrower shall have the 4.01(b) The Government sent IDA the accounts for each fiscal year project audit report, due by audited. the end of June 1986, in August 1987. The Borrower shall consult 4.03 Not complied with. with the Association on its policy to develop the Djiboutian road sector and the investments there in. The Borrower shall take all 4.04 Not complied with. actions as may be necessary to maintain adequately the National Road Network.  - 23 - 3.07 Use of Bank Resources A. Staff Inputs in staff-weeks Stage of Project Cycle Planned Revised Final Comments Through Appraisal - - 29.1 Appraisal through Effectiveness - - 7.6 Supervision - - 24.2 Total - - 60.9  - 24 - 3.07 Use of Bank Resources (continued) B. MISSIONS Number Days Performance Types Stage of Month/ of in Specialization Ratinz of Pioject Cycle Year Persons Field Represented 1 Status Problems Throu,h Appraisal 10/80 1 4 1 EGR 11/81 3 14 1 ECO/ 2EGR AppraLsal through Board Approval - - - ?-ard Approval .Lrou,h Effec:iveness 01/83 2 5 1 ECO / 1 EGR - Super ,ision 12/83 1 4 1 EGR 1 07/84 1 5 1 EGR 1 - 05/85 1 8 1 EGR 1 - 01/86 1 7 1 EGR 1 - 10/86 2 8 2 EGR 1 - 1. Kiy to Specialization: E:O - Economist EIR - Engineer T'O = Training Officer 2. Kiy to Status: 1 = Problem free or minor problems 2 = Moderate problems 3 = Major problems  - 25 - Annex 1 Page 1 of 2 DJIBOUTI HIGHWAY MAINTENANCE PROJECT Credit 1303-DJI Traffic and VOC I. Traffic 1. Traffic on Unpaved Roads Road Type Cars Pick-ups 4-wd Trucks Total vpd % vpd % vpd % vpd % vpd % 20 VpD 5 25 2 10 4 20 9 45 20 100 10 VpD 3 30 1 10 2 20 4 40 10 100 5 VpD 1 20 1 20 1 20 2 40 5 100 2. Traffic on Paved Roads Road Type Cars Pick-ups 4-wd Trucks Total vpd % vpd % vpd % vpd % vpd % National Roads 15 25 6 10 12 20 27 45 60 100 Streets 27 30 9 10 18 20 36 40 90 100 II. VOC 1. VOC in DF/Veh.km for Unpaved Roads1_ Vehicle Type Surface Type Gravel Earth Cars 25 89 Pick-Ups 40 89 4-WD 53 89 Trucks 75 110 Average 54 98 1 These are the only vehicle operating costs available on file (see "Etude Diagnostique de L'Entretien Routier, BCEOM").  - 26 - Annex 1 Page 2 of 2 VOC in DF/Veh.km for Paved Roads The following basic VOC for paved roads, because not available on file, were ,ved from those for unpaved roads shown in para II-1 here above and from VOC other countries with similar climatic and topographical conditions as outi. Vehicle Type Pavement Condition Very Good Fair Bad Bad Cars 23 25 36 45 Pick-Ups 37 40 57 70 4-WD 49 53 76 94 Trucks 63 76 100 134 Average 47 54 74 96  - 27 - Annex 2 Page 1 of 3 DJIBOUTI HIGHWAY MAINTENANCE PROJECT Credit 1303-DJI Economic Evaluations Paved Roads 1. Economic Re-evaluationL of Resurfacing RN 1 Resurfacing Maintenance VOC Years Cost Savings Savings Cash Flow ERR --------------------- DFx1000 ----------------------- % 1985 -36,947 -36,947 16 1986 -34,105 -34,105 1987 3,651 7,117 10,768 1988 4,637 8,503 13,140 1989 5,889 10,347 16,236 1990 7,479 12,735 20,214 1991 7,918 15,578 23,496 1992 10,150 18,728 28,878 1993 12,773 22,763 35,536 2. Economic Evaluation of Resurfacing RN 3 and 5 Resurfacing Maintenance VOC Years Cost Savings Savings Cash Flow ERR -_______________----- DFxl000 ----------------------- Z 1986 -95,920 -95,920 10 1987 3,213 6;,263 9,476 1988 4,081 7,162 11,243 1989 4,710 8,227 12,937 1990 5,436 9,479 14,915 1991 3,061 10,855 13,916 1992 3,621 12,261 15,882 1993 4,278 13,901 18,179 1994 5,049 15,875 20,924 1995 5,953 18,236 24,189 1996 7,012 21,126 28,138  - 28 - Annex 2 Page 2 of 3 3. Economic Evaluation of Resurfacing Two Streets of Diibouti Improvement Maintenance VOC Years Cost Savings Savings Cash Flow ERR - ------------------- DFxl000 ---------------------- Z 1985 -14,222 -14,222 25 1986 -68,554 -68,554 1987 3,359 9,822 13,181 1988 4,266 11,374 16,000 1989 5,418 14,278 19,696 1990 6,881 17,574 24,455 1991 5,380 21,498 26,878 1992 9,338 25,844 35,182 1993 11,752 31,412 43,164 4. Economic Re-Evaluation of Routine Maintenance Avoided Improvement Cost VOC Years Cost of Repairs Savings Cash Flow ERR ----------------------- DFxl000 ---------------------- % 1984 -104,411 -104,411 57 1985 -104,411 45,990 -58,421 1986 -104,411 95,856 -8,555 1987 407,203 152,015 559,218  - 29 - Annex 2 Page 3 of 3 Earth Roads 5. Economic Evaluation of Track Rehabilitation Improvement Maintenance VOC Years Cost Savings Savings Cash Flow ERR ---------------------- DFxl000 --------------------- % 1985 -129,000 -129,000 14 1986 -133,000 -133,000 1987 30,668 13,651 44,319 1988 30,668 13,651 44,319 1989 30,668 13,651 44,319 1990 30,668 13,651 44,319 1991 30,668 13,651 44,319 1992 30,668 13,651 44,319 1993 30,668 13,651 44,319 1994 30,668 13,651 44,319 1995 30,668 13,651 44,319 1996 -24,072 13,651 -10,421 1997 -24,072 13,651 -10,421 1998 30,668 13,651 44,319 1999 30,668 13,651 44,319 2000 30,668 13,651 44,319 2001 30,668 13,651 44,319 2002 30,668 13,651 44,319 2003 30,668 13,651 44,319 2004 30,668 13,651 44,319 2005 30,668 13,651 44,319 2006 30,668 13,651 44,319 2007 128,000 30,668 13,651 172,319  MAP SECTION  DJIBOUTI HIGHWAY MAINTENANCE PROJECT . ~~~~~~~~~~. ,T8~, - ..~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ . ., _ PAN,, C, / x - + te1- f \ bjz tt10 sq S,~~~~~~~~~~~~~~~~~~~~~~~~S //OV ETHIOPIAJ .. si~~~~~~~~~~~~~~~~~